Cancer occurrence in Danish diabetes patients: Duration and insulin effects

Transcription

1 Cancer occurrence in Danish diabetes patients: Duration and insulin effects Compiled Thursday 20 th October, 2011, 16:12 from: C:/Bendix/Steno/DMregister/NDR/projects/Cancer/papers/P1/Updated2009DMCA4.tex Word count: 3877 Bendix Carstensen Corresponding author Danel R Witte Søren riis Steno Diabetes Center, Gentofte, Denmark & Department of Biostatistics, University of Copenhagen bxc@steno.dk Conflict of interest: Stockholder of NovoNordisk Steno Diabetes Center, Gentofte, Denmark drw@steno.dk Conflict of interest: Stockholder of NovoNordisk Institute of Cancer Epidemiology Danish Cancer Society, Copenhagen Denmark & Institute of Public Health, University of Copenhagen friis@cancer.dk

2 Abstract Background: Cancer is more frequent among diabetes patients, but it is unknown how this excess varies with duration of diabetes and insulin use. Methods: We linked the Danish National Diabetes Register and Cancer Registry and performed a cohort analysis of the entire Danish population by diabetes status, duration of diabetes and insulin use, comparing cancer incidence rates in diabetes patients to the non-diabetic population, using Poisson regression to describe the variation by duration. Results: We found 20,032 cancer cases among patients not using insulin and 2,794 cancer cases among diabetes patients using insulin. The cancer incidence rate-ratio among non-insulin users relative to the non-diabetic population decreased from over 2 at diagnosis to 1.15 after 2 years of diabetes duration. The cancer incidence rate-ratio was higher among patients using insulin, decreasing from 5 at start of insulin to about 1.25 after 5 years of insulin use. Long-term elevated risk was seen among non-insulin users for cancers of colorectum, corpus uteri, kidney, liver and pancreas; among insulin users for cancers of the stomach, lung, kidney and liver. Conclusions: The observed duration effects suggest that both increased surveillance for cancer in the first years after diagnosis of diabetes, as well as reverse causation where undiagnosed cancers increase the likelihood of diabetes diagnosis play a role. or longer durations, a combination of common causes for diabetes and cancer, as well as effects of diabetes and insulin exposure per se may play a role in the association between diabetes and some cancers. 0

3 Abbreviations: CPR: Central Person Register DCR: Danish Cancer Registry DM: Diabetes Mellitus ESM: Electronic supplementary material NDR: (Danish) National Diabetes Register OAD: Oral antidiabetic drugs RR: Rate ratio Keywords: Diabetes duration, insulin treatment, insulin duration, cancer incidence, cancer rate-ratio, mortality, register study, prescriptions. 1

4 2 0 Introduction Introduction Diabetes patients carry a higher risk of cancer than the general population and this excess risk differs by cancer site [1, 2, 3]. The explanation has remained elusive as the low incidence of any given cancer means that most studies in the field have been observational by necessity and therefore open to confounding that cannot be remedied. A number of studies have addressed the question of the relationship between drug use and cancer occurrence among diabetes patients, particularly with respect to the effect of oral anti-diabetic drugs (OADs) [4]. A study of the cancer incidence in insulin users found a 2-fold increased risk of colorectal cancer among insulin users. Several studies have reported an inverse association between diabetes and subsequent prostate cancer [5, 6]. However, the nature of the link between the risk of cancer and the duration of diabetes and insulin use is not well described, let alone understood. The aim of this study was to relate the occurrence of cancer among diabetes patients to the non-diabetic part of the population and in particular to describe how this relation varies by duration of diabetes and insulin treatment. Material and methods Register data The study covered the entire Danish population in the 15 year period from to , based on register linkage. We used registers to retrieve the date of birth, date of cancer, date of diabetes, date of first insulin use and date of death for all persons in the population. Central Person Register (CPR) Since , all persons with permanent residence in Denmark have been given a unique identification number (CPR-number) [7] which is used for all administrative purposes in Denmark, in particular, all health events recorded in registers are identified by the CPR-number, and so are uniquely linkable.

5 Cancer among Danish diabetes patients 3 The National Diabetes Register (NDR) The National Diabetes Register was established in 2006, and currently contains records of all prevalent cases of diabetes as of and all incident cases of diabetes until [8, 9]. The start of insulin treatment is defined as the date of the second insulin purchase in the prescription database. The recorded date of inclusion in the NDR is only reliable after Therefore, duration of diabetes and insulin use can only be determined reliably for persons included in the register after The Danish Cancer Registry (DCR) The Danish Cancer Registry was established in 1943 [10, 11], and contains all tumours diagnosed in Denmark between and , a total of 1.25 million cancers among 1.15 million persons. All patients alive after , when the CPR was established, are identified by their CPR number. Cancers in this study were defined as the first primary cancer only, in order to avoid diagnostic artifacts and possible effects from treatment of the first cancer. Moreover, non-melanoma skin cancer was not counted as a cancer in this study. Linkage and tabulation Persons were followed from to , death or diagnosis of any cancer. All prevalent cases of cancer and diabetes as of were excluded. ollow-up was classified in the categories no DM, DM (DM, no insulin) and DM+Ins (DM, insulin treatment). Person-years and the number of incident cancers were enumerated for 25 different groups of cancer diagnoses, as were deaths prior to cancer diagnosis (Table 1). Cancers and follow-up during the first month after diabetes diagnosis were excluded to avoid uncertainty of the sequence of diabetes and cancer diagnoses (details in the Electronic Supplementary Material (ESM)). The person-years in the Danish population were based on figures from Statistics Denmark and calculated as indicated in [12, 13, 14]. The person-years in the no DM

6 4 Statistical model and analysis category were computed as the person-years in the total Danish population minus the person-years lived by persons diagnosed with cancer or diabetes. The follow-up (person-years and cancer cases) was classified by diabetes and treatment category, sex, age and calendar time at follow-up and date of birth in 1-year classes (Lexis triangles, [12]). ollow-up in the two diabetic categories was further subdivided by time since inclusion in the register (diabetes duration), and by time since second insulin purchase (insulin duration), in intervals of length 3, 6, 12 and 24 months, as illustrated in figure 2 in the ESM. All data manipulation and tabulation was performed in the SAS system, using the macro %Lexis, [15]. This is documented in detail in the links provided in the ESM. Statistical model and analysis We fitted multiplicative Poisson models for the rates (proportional hazards models), separately for each sex and site of cancer and for all cancers combined. As continuous covariates we used age, calendar time and date of birth midpoints for each cell in the table [12], as well as the duration of diabetes / insulin treatment at the midpoint of each duration interval. Model for the rate-ratio or the non-diabetic part of the population, we used an age-period-cohort (APC) model for the cancer incidence rates, using cubic splines with a total of 20 parameters [12]. irst a simple model was fitted assuming no duration effects, but only a fixed rate-ratio (RR) for each of the categories DM and DM+Ins ; mainly to preserve comparability with previously published studies. We tested whether insulin users had the same cancer risk as patients not using insulin, and whether diabetes patients had the same risk as the non-diabetic population. In an extended model, the RR relative to the non-diabetic population was taken as a function of diabetes duration and (for insulin users), insulin duration, modeled by restricted cubic (natural) splines with 4 and 3 parameters, respectively (for the outcome all cancer, 5 and 4 parameters) [16]. We chose the first knot to be 0 and the remaining

7 Cancer among Danish diabetes patients 5 knots so that the number of incident cancers between each pair of knots and after the last was the same. Interaction between insulin and diabetes duration was tested by including two extra terms: diabetes duration at insulin start and the product of diabetes and insulin duration. The RR is shown graphically by duration of diabetes for diabetes patients not on insulin, as well as for diabetes patients initiating insulin at 0, 3 and 5 years after diagnosis. Cumulative risk To illustrate how the estimated rates and RRs translate into risk of cancer, we computed the 10-year cumulative risk of each particular type of cancer from age 65, 70 or 75 for a person in each of the categories No DM, DM and DM+Ins. The follow-up was assumed to cover the period to We accounted for death as a competing risk to cancer [17, 18, 19] by using mortality rates estimated from APC-models for death fitted separately for the categories No DM, DM and DM+Ins Terminology We used the term significant for estimates that were more than twice as large as the estimated standard error, even though the formal statistical machinery relies heavily on the correctness of the models used. Likewise, the 95% confidence limits used were based on the correctness of the models, and the validity of the normal approximation to the distribution of the parameter estimates. All statistical analyses and graphical reporting of results were made in R [20]. These are documented in detail in the material in the ESM links..

8 6 0 Results Results igure 1 shows the overall RRs of cancer relative to the non-diabetic population for persons with diabetes subdivided by insulin use. The RRs among diabetes patients on insulin and those not on insulin followed largely the same pattern with respect to site of cancer, with a general tendency toward higher RRs among those on insulin. or all cancers combined, the RR was 1.2 for patients not on insulin and 1.4 for patients on insulin, with very slight differences between the sexes. In the non-insulin group, there were significantly elevated RRs for cancer of the stomach, colorectum, liver, pancreas, lung, corpus uteri, kidney, brain and lymphomas, whereas elevated rates among insulin users were seen for stomach, liver, pancreas, lung, corpus uteri and kidney. The RRs in the insulin group were generally higher, but fewer were significant. The RR for liver cancer was 3.4 among males and 1.6 among females not on insulin, but more than twice as high among those on insulin (8.1 and 4.0, respectively). or pancreatic cancer, the RR was 2.4(M) and 2.2() for patients not on insulin, and three times as high (6.9 and 7.2) among those on insulin. The RR for prostate cancer among patients on insulin was 0.8, whereas that for patients not on insulin was 1.0. We found a decreasing tendency for the RR with more distant locations along the digestive tract with the highest RR for stomach cancer and smallest for rectal cancer, see figure 1. Within the group of diabetes patients, we found significantly higher cancer incidence rates among insulin users relative to non-users for all cancers combined, and for cancers of the liver, pancreas, lung and kidney, and smaller rates of prostate cancer, see figure 1 and the figure of RRs comparing users and non-users shown in the ESM. No other sites exhibited a consistent pattern in the overall effect of DM or insulin treatment on cancer risk.

9 Cancer among Danish diabetes patients 7 Duration effects Among the 25 sites analysed by sex, we only found a significant interaction between duration of diabetes and insulin for two sites (liver, M and thyroid, ). Moreover, including the estimated interaction in the reporting of the effects makes virtually no difference in the curves (data not shown). Hence the model without interaction between the durations was used for reporting throughout. The general shape of the RR curve by duration showed a high risk at the beginning, decreasing during the first year to a fairly moderate level. The shapes of the RR curves were quite similar for insulin users and non-users (figure 2). or all cancers combined, the RR among non-users was highest in the period just after diagnosis with a RR of 2.0 decreasing to 1.15 after two years. or insulin users the RR started at 5, decreasing to 1.3 after about 3 years. Among males, there was a slight tendency for RRs to decline for longer insulin duration. The RRs were statistically significant over the entire study period, with the exception of male insulin users where the effect was not significant after 11 years. or all cancers combined, the RR between insulin users and non-users was about 2.5 at initiation of insulin, about 1.2 after 4 years of insulin use, and largely constant for females, but slightly declining for males, see figure 3. These RRs were significant for the first 7 9 years of insulin duration. The effects of duration of diabetes and insulin therapy are shown in figures 4 and 5 for the cancer sites stomach, colorectal, lung, prostate, breast, ovary, corpus uteri, kidney, urinary bladder, liver, pancreas and melanoma of skin. These are the sites with the largest number of cancer cases among diabetes patients similar figures for all analysed sites are included in the ESM. Stomach cancer showed a constant RR of about 1.2 for the non-insulin treated diabetes patients (non-significant when modeled by duration), and an indication of a strong duration effect among insulin-using women only. Colorectal cancer showed long-term elevated RR of about 1.2 among non-insulin treated patients, and no indication of a difference between insulin-users and non-users. or lung cancer, we did not find a long-term elevated RR among non-insulin users, but among insulin users the long-term RR was about 1.4 for both sexes.

10 8 Cumulative risk or melanoma of skin, we saw a tendency toward RR below 1 with increasing duration among women, both for insulin users and non-users, whereas there was no indication of an effect in either direction for males. There was no allocation effect (initially increased RR) for melanoma of skin in either sex. The smaller incidence rates of prostate cancer among insulin treated patients were accentuated by longer diabetes duration or insulin treatment. There was no indication of a breast cancer RR different from 1 in any direction regardless of insulin treatment or duration. Cancer of the corpus uteri showed an RR of around 1.6 at all durations with no indication of any duration effects apart from a small initial allocation effect. Liver cancer showed duration patterns for users and non-users of insulin which were consistent with a largely constant RR. Pancreas cancer showed a dramatic decrease in RR by both diabetes and insulin duration. Kidney cancer showed an initial decrease in RRs for both sexes, and for males, we found a long term constant RR of around 1.3 for non-users and 1.7 for users of insulin (not significantly different), whereas there were higher RRs among females, and even an indication of an increasing RR with longer duration of diabetes. The RR between users and non-users of insulin was constant around 2 and significant for the first 8 years of insulin use. No consistent patterns by diabetes duration and insulin duration were seen for cancers of the ovary or urinary bladder. Cumulative risk The 10-year cumulative risks of cancer or dying without cancer are shown in figure 6. The differences in cancer incidence rates between the groups were largely dominated by the mortality differences, and the cumulative risk of cancer was approximately the same for the three groups. As a result the 10-year cumulative risks of cancer essentially only varied by sex (M: 17 18%; : 14 16%, for 65 year olds). Table 1 presents site-specific cumulative cancer risks for each of the cancer sites from age 65 to 75 years, assuming that follow-up began in 1998 at diabetes diagnosis and insulin start. The electronic supplementary material contains graphs of RRs for all sites analysed. A complete account of all analyses reported, including SAS and R-code can be found at

11 Cancer among Danish diabetes patients 9 the first author s website as: Discussion In this study, we compared the cancer occurrence in newly diagnosed Danish diabetes patients with the non-diabetic part of the population, and we compared patients on insulin with patients not on insulin. or the first time in the literature, we have simultaneously modeled both inception and duration effects of diabetes and insulin treatment. We found that the risk of cancer decreased substantially for most sites by duration of diabetes or insulin use. The elevated cancer risk observed just after diabetes diagnosis and the marked decrease during the first two years were seen for almost all sites of cancer. This indicates presence of a detection bias, in the sense that the diagnosis of diabetes leads to increased medical attention, and thus to earlier detection of any present but undiagnosed cancer. The large number of patients with dates of diagnosis of cancer and diabetes very close to each other (excluded from analysis because of lack of evidence that diabetes preceded cancer see the ESM), also suggests that more or less concomitant clinical manifestation or registration of cancer and diabetes may be quite frequent. The stable level of the rate ratio seen beyond the initial 2 year period may reflect the combined effect of risk factors common to diabetes and cancer, such as obesity and lifestyle factors, but also pre-diabetic conditions characterized by moderately elevated blood glucose and insulin resistance. Stocks [21] found that there was an effect of higher blood glucose levels on the occurrence of cancers, particularly for higher levels of blood glucose with a pattern of elevated RRs across cancer sites not very different from the one seen in this study. Moreover, OADs may have an effect on cancer risk, as indicated by the higher cancer rates seen for patients using sulphonylureas relative to those using metformin [4].

12 10 Insulin effects Insulin effects We observed the largest RR between the insulin treated and the non-diabetic population at the time of initiation of insulin treatment, with a sharp drop in RR during the first year, regardless of the time of insulin initiation relative to diabetes onset (lack of interaction). This suggest an indication bias: As insulin is generally the third or fourth treatment option, the clinical decision to initiate treatment depends on failure of earlier treatment, or contra-indications to OADs, such as heart failure, renal or hepatic conditions which are indicative of general ill health. To the extent that this is linked to undiagnosed (or emerging) cancers, a high RR can be expected in people with a recent start of insulin treatment. Moreover, detection bias may also play a role, where people with a recent insulin prescription are checked more thoroughly and monitored more closely. or all cancer sites combined, we observed a largely constant RR (1.3) for insulin-treated versus the non-diabetic population with longer duration of insulin use, at least during the first 10 years. A direct effect of insulin on cancer initiation or promotion cannot be ruled out. In isolation such an effect would be expected to start at RR=1 and increase continuously by duration. We observed an initial RR in insulin treated versus non-insulin treated diabetes patients of about 2.0 which fell to a level of 1.2 after 2 years among women; slightly less among men. This suggests that other causes than a direct insulin effect play a dominant role just after treatment initiation, but a small slowly increasing effect of insulin use cannot be ruled out. In the assessment of potential insulin effects, the all cancers category is not really biologically meaningful. In this study, prostate, melanoma and colorectal cancer followed the pattern of decreasing RR with longer insulin duration described above, whereas cancers of the stomach, lung, liver and kidney showed an indication of elevated cancer risk among insulin users. Most strikingly, the RR for pancreatic cancer showed a continued very sharp decline with increasing insulin duration. This effect is very likely due to reverse causation, where a pre-existing sub-clinical pancreatic cancer is the cause of diabetes, and first

13 Cancer among Danish diabetes patients 11 diagnosed after diabetes diagnosis. We found no indication that breast cancer incidence among diabetes patients differed from the non-diabetic population regardless of duration of diabetes or insulin use. This is in line with the findings of Johnson et al. [22], and studies not taking duration into account [1, 2, 3], whereas Coughlin et al. [23] found an elevated risk of breast cancer in a mortality study. Relation to other studies Johnson et al. [22] recently analysed the effect of diabetes duration on cancer incidence in British Columbia, Canada, and found broadly similar results to ours, with elevated cancer incidence RR in the first months after diagnosis particularly for colorectal, prostate, lung and pancreatic cancers. They found long-term elevated RR only for cancer for the corpus uteri (endometrium) (RR=1.6), and a reduced RR of 0.8 for prostate cancer, consistent with our long term estimates. Similarly they also found that breast cancer rates were not affected by diabetes status or duration. Our study also uniquely included information on insulin usage, as well as duration of diabetes and insulin treatment, and hence represents an essential addition to four recently published studies aiming at studying effects of specific insulin analogs [24, 25, 26, 27]. These four studies only performed internal comparisons between selected groups of diabetes patients and did not relate the cancer occurrence to that in the general population, because they focused on particular drug combinations. Moreover, these studies were largely confined to a short term follow-up, i.e. the time window where our study indicates that substantial confounding is present. Our study specifically aimed at studying duration effects, and hence only included persons diagnosed after Previous studies [1, 2, 3, 23] also included prevalent diabetes cases, and thus had an average length of diabetes duration and insulin duration that was longer than that in our study. Since the risk is highest for short duration of diabetes, we should expect to see higher RRs in our study s overall analysis (figure 1). or comparison, we did a simple analysis ignoring both diabetes duration and insulin treatment (see figure 2.13 in the ESM with a comparison to previous major studies), and indeed found that this approach yielded RRs which were generally higher than in

14 12 Cumulative risks previous studies that ignored diabetes duration, but included prevalent cases of DM in the follow-up. It could therefore be argued that analyses ignoring diabetes duration are meaningless as they average over an influential variable (duration) with weights which are strongly dependent on the study design. uture studies of cancer occurrence among diabetes patients should in our view always include diabetes duration in order to distinguish between the markedly different risk periods we (and Johnson et al.[22]) detected. The NDR does not distinguish type 1 and type 2 diabetes, but a tabulation of cancers among patients diagnosed under age 35 showed a total of 11 cases, so type 1 patients contribute minimally to the results. Cumulative risks Our study is the first to document that the significant differences in cancer incidence between diabetes patients with or without insulin therapy and non-diabetics does not translate into major differences in the cumulative risk of cancer. Over a 10-year period, a diabetes diagnosis and initiation of insulin treatment were much stronger predictors of death from other causes than cancer. Due to the limited 10-year survival probability for diabetes patients, the cumulative risk of a cancer diagnosis was similar to that of among non-diabetics. rom a public health point of view, mortality from other causes than cancer seems to be a greater health issue among diabetes patients than the potential excess in cancer risk associated with diabetes and insulin therapy. Limitations Our study falls a bit short of recently published smaller studies, because the only available information on medication use was the date of second insulin purchase and no information on different insulin types or dosage was available. As all observational studies of effects of diagnosis and treatment, the current study is subject to confounding by indication, and the observed RRs between DM patients and insulin treated relative

15 Cancer among Danish diabetes patients 13 to the general population cannot be interpreted causally. Strengths The study covered the entire Danish population, and thus there was no selection bias with regard to entry into the study population. Moreover, the study gives a precise picture of the actual incident cancer burden, as the Danish Cancer Registry s long history allowed for effective exclusion of all prevalent cancer cases at the start of our follow-up period. By including duration in our analyses, we were able to correct for the spurious effects seen in the first period after diagnosis of diabetes, and thus provide more credible estimates of the effects of diabetes and insulin use. inally, the thorough Danish registration system ensured that all emigrations and deaths in the study population were properly accounted for. Conclusions Diabetes patients have elevated cancer incidence rates compared to the non-diabetic population, highest in the first year after diagnosis and the first year after initiation of insulin. or non-users of insulin, the long-term RR was 1.1 and for insulin users the long-term RR was 1.3. Diabetes patients not using insulin have higher long-term incidence rates of cancer of the colorectum, corpus uteri, kidney, liver and pancreas. Diabetes patients using insulin have higher long-term incidence rates of cancer of the stomach, lung, kidney and liver. emale diabetes patients have the same breast cancer incidence rates as those without diabetes, regardless of insulin therapy and duration. Male diabetes patients have decreased rates of prostate cancer, the rate ratio decreases by duration of diabetes and particularly with the duration of insulin use. Clinical effects of diabetes diagnosis and start of insulin treatment are clearly present, but long term effects of diabetes per se as well as insulin cannot be ruled out.

16 14 0 Disclosure Disclosure Bendix Carstensen and Daniel R. Witte are stockholders of NovoNordisk and employees at Steno Diabetes Center, a research hospital owned by NovoNordisk. Søren riis has no conflicts of interest in relation to this work. An ethics statement was not required for this work. No funding was received for this work, no funding bodies played any role in the design, writing or decision to publish this manuscript. The study was conceived collectively by the three authors, data preparation and statistical analysis was done by BC, who also wrote the first draft of the paper. All three authors contributed equally to critical revision and interpretation of findings, and have all approved the final version of the paper. A preliminary version of this work, based only on 14 years of follow up (not 15), however, with an unfortunate computing error, was presented as a paper at the EASD meeting in September 2010 in Stockholm. In the abstract book (Abstract 147), a graph showed overall RRs of more than 1.5 associated with long term use of insulin. The correct figure is 1.3 as presented in this paper, figure 2.

17 Cancer among Danish diabetes patients 15 References [1] Adami HO, McLaughlin J, Ekbom A, Berne C, Silverman D, Hacker D, Persson I. Cancer risk in patients with diabetes mellitus. Cancer Causes Control, 2: , Sep [2] La Vecchia C, Negri E, ranceschi S, D Avanzo B, Boyle P. A case-control study of diabetes mellitus and cancer risk. Br. J. Cancer, 70: , Nov [3] Wideroff L, Gridley G, Mellemkjær L et al. Cancer incidence in a population-based cohort of patients hospitalized with diabetes mellitus in Denmark. J. Natl. Cancer Inst., 89: , Sep [4] Bowker SL, Majumdar SR, Veugelers P, Johnson JA. Increased cancer-related mortality for patients with type 2 diabetes who use sulfonylureas or insulin. Diabetes Care 29: , 2006, 29: , [5] Bonovas S, ilioussi K, Tsantes A. Diabetes mellitus and risk of prostate cancer: a meta-analysis. Diabetologia, 47: , Jun [6] Calton BA, Chang SC, Wright ME, et al. History of diabetes mellitus and subsequent prostate cancer risk in the NIH-AARP Diet and Health Study. Cancer Causes Control, 18: , Jun [7] Pedersen CB. The Danish civil registration system. Scand J Public Health, 39 (7 suppl):22 25, July [8] Carstensen B, Kristensen JK, Ottosen P, Borch-Johnsen K. The Danish National Diabetes Register: Trends in incidence, prevalence and mortality. Diabetologia, 51: , [9] Carstensen B, Christensen JK, Marcussen MM, Borch-Johnsen K. The National Diabetes Register. Scandinavian Journal of Public Health, 39(7 suppl):58 61, [10] Storm HH, Michelsen EV, Clemmensen IH, Pihl J. The Danish Cancer Registry history, content, quality and use. Dan Med Bull, 44: , Nov [11] Gjerstorff ML. The Danish cancer registry. Scand J Public Health, 39/7 Suppl):42 45, July [12] Carstensen B. Age-Period-Cohort models for the Lexis diagram. Statistics in Medicine, 26(15): , July [13] Rosenbauer J, Strassburger K. Comments on: Age-Period-Cohort models for the Lexis diagram. Statistics in Medicine, 27: , [14] Carstensen B. Age-Period-Cohort models for the Lexis diagram (author s reply). Statistics in Medicine, 27: , 2007.

18 16 REERENCES [15] Carstensen B. %Lexis: a SAS-macro for splitting follow-up time. Accessed Thursday 20 th October, [16] Harrell E. Regression modeling strategies. Springer, [17] Putter H, iocco M, Geskus RB. Tutorial in biostatistics: competing risks and multi-state models. Stat Med, 26: , May [18] Andersen PK, Abildstrøm SZ, Rosthøj S. Competing risks as a multi-state model. Statistical Methods in Medical Research, 11: , [19] Cornfield J. Estimation of the probability of developing a disease in the presence of competing risks. Am J Public Health Nations Health, 47(5): , [20] R Development Core Team. R: A Language and Environment for Statistical Computing. R oundation for Statistical Computing, Vienna, Austria, ISBN [21] Stocks T, Rapp K, Bjørge T, et al. Blood glucose and risk of incident and fatal cancer in the metabolic syndrome and cancer project (Me-Can): analysis of six prospective cohorts. PLoS Med., 6:e , Dec [22] Johnson JA, Bowker SL, Richardson K, Marra CA. Time-varying incidence of cancer after the onset of type 2 diabetes: evidence of potential detection bias. Diabetologia, 54: , Sep [23] Coughlin SS, Calle EE, Teras LR, Petrelli J, Thun MJ. Diabetes mellitus as a predictor of cancer mortality in a large cohort of US adults. Am. J. Epidemiol., 159: , Jun [24] Jonasson JM, Ljung R, Talbäck M, Haglund B, Gudbjörnsdóttir S, Steineck G. Insulin glargine use and short-term incidence of malignancies-a population-based follow-up study in Sweden. Diabetologia, 52: , Sep [25] Colhoun MH and the SDRN Epidemiology Group. Use of insulin glargine and cancer incidence in Scotland: a study from the Scottish Diabetes Research Network Epidemiology Group. Diabetologia, 52: , Sep [26] Currie CJ, Poole CD, Gale EA. The influence of glucose-lowering therapies on cancer risk in type 2 diabetes. Diabetologia, 52: , Sep [27] Hemkens LG, Grouven U, Bender R, Günster C, Gutschmidt S, Selke GW, Sawicki PT. Risk of malignancies in patients with diabetes treated with human insulin or insulin analogues: a cohort study. Diabetologia, 52: , Sep 2009.

19 Table 1: Person-years and cancer cases in diabetes free persons and in diabetes patients diagnosed in the 15-year study period incl. Only first primary cancers are listed. Non-melanoma skin cancer is not counted as a cancer diagnosis. The null hypotheses tested are that the RR is 1 at any duration. No. cancer cases Cumulative risk age 65 75(%) P-value for null: No DM 4 DM DM+ins No DM 4 DM DM+ins Ins. = DM DM = Pop. M M M M M M M M Person-years (1000s) 37, , All malignant neoplasms 167, ,744 11,378 8,654 1,609 1, Oesophagus 3,282 1, Stomach 4,171 2, Colorectal cancer 1 22,729 21,350 1,725 1, Ascending colon 3,636 5, Transverse colon 1,819 2, Descending colon 7,292 6, Rectum 8,876 6, Liver 1,776 1, Pancreas 4,041 4, Lung, bronchus and pleura 25,302 20,283 1,673 1, Melanoma of skin 7,008 8, Breast , , Cervix uteri 5, Corpus Uteri 8, Ovary 7, Prostate 31,713 2, Testis 4, Kidney 3,976 2, Urinary bladder 14,781 4,906 1, Brain 4,922 4, Thyroid 593 1, Hodgkin s lymphoma Non-Hodgkin lymphoma 5,434 4, Multiple myeloma 2,452 1, Leukaemia 5,423 4, Other 2 24,246 21,360 1,502 1, Deaths 3 225, ,814 19,737 17,987 3,872 2, Includes unspecified colon cancers in addition to the four sub-sites listed in the table. 2 Cancers not in any of the mentioned groups. Not analysed. 3 Only deaths occurring before first primary cancer. 4 irst primary cancers among persons without diabetes. Cancers among prevalent cases of diabetes as of are not in this table.

20 igures All malignant neoplasms Oesophagus Stomach Colorectal cancer Ascending colon Transverse colon Descending and sigmoid colon Rectum Liver Pancreas Lung, bronchus and pleura Melanoma of skin Breast Cervix uteri Corpus uteri Ovary, fallopian tube etc. Prostate Testis Kidney Urinary bladder Brain Thyroid Hodgkin's lymphoma Non Hodgkin lymphoma Multiple myeloma Leukaemia Light color: insulin users M RR, DM / DM+ins vs. non DM igure 1: Estimated rate-ratios of cancer occurrence for DM patients not on insulin (full color) and DM-patients on insulin (light color) versus non-dm-patients for different cancers. Males: blue; females: red. The RRs for cancer of the liver, pancreas and testis in insulin treated patients are off the figure: Liver, M: 8.1 (6.6;9.9) : 4.0 (2.6;6.3); Pancreas: M: 6.9 (6.0;7.9) : 7.2 (6.2;8.4); Testis: 0.50 (0.24;1.05).

21 5.0 All malignant neoplasms M Rate ratio: DM/no Ins, DM/Ins vs. non DM Diabetes duration (years) igure 2: Cancer incidence rate-ratio versus the non-diabetic population. Spline models with 5 (DM-duration) and 4 (insulin duration) parameters. The full lines are for DMpatients not on insulin, the broken lines are for patients on insulin, starting insulin after 0, 2 and 5 years of DM duration, respectively. Thin lines indicate 95% confidence intervals. The bars on the left are estimates from the model ignoring duration effects, and the coloured ticks inside indicate the location of the knots for the spline functions the number of cancers between each pair of knots and to the right of the last are the same.

22 5.0 All malignant neoplasms M Rate ratio: DM/Ins vs. DM/no Ins Insulin duration (years) igure 3: Cancer incidence rate-ratio for insulin users versus non users. Spline models with 5 (DM-duration) and 4 (insulin duration) parameters. Thin lines indicate 95% confidence intervals. The bars on the left are estimates from the model ignoring duration effects, and the coloured ticks inside indicate the location of the knots for the spline functions.

23 5.0 Stomach M 5.0 Colorectal M Rate ratio versus the non diabetic population Lung Prostate M M Breast Melanoma Ovary M Corpus uteri Diabetes duration (years) igure 4: Rate-ratio for selected sub-sites versus the non-diabetic population. Results from spline models with 4 (DM-duration) and 3 (insulin duration) parameters. The full lines are for DM-patients not on insulin, the broken lines are for patients on insulin, starting insulin after 0, 2 and 5 years of DM duration, respectively. Thin lines indicate 95% confidence intervals. The bars on the left are estimates from the model ignoring duration effects, and the coloured ticks inside indicate the location of the knots for the spline functions the number of cancers between each pair of knots and to the right of the last are the same.

24 Liver M Pancreas M Rate ratio versus the non diabetic population Kidney M Urinary bladder M Diabetes duration (years) igure 5: Rate-ratio for selected sub-sites versus the non-diabetic population. Results from spline models with 4 (DM-duration) and 3 (insulin duration) parameters. The full lines are for DM-patients not on insulin, the broken lines are for patients on insulin, starting insulin after 0, 2 and 5 years of DM duration, respectively. Thin lines indicate 95% confidence intervals. The bars on the left are estimates from the model ignoring duration effects, and the coloured ticks inside indicate the location of the knots for the spline functions the number of cancers between each pair of knots and to the right of the last are the same.

25 100 Age at start: 60 Age at start: 65 Age at start: year cumulative risk of Cancer / Death M M M No DM DM DM+Ins M M M No DM DM DM+Ins M M M No DM DM DM+Ins igure 6: 10-year cumulative risk for any type of cancer (light bottom part) and death (full colored middle part) and survival (pale color, top part) from ages 60, 65 and 70, starting A non-diabetic person is assumed immune to diabetes and a patient not on insulin is assumed to remain free from insulin.